Oracle Exalogic Elastic Cloud Software
 
Enterprise Manager Ops Center
 
Exalogic Elastic Cloud X5-2 Eighth Rack
 

PLA-Support>Eng Systems>Exalogic/OVCA>Oracle Exalogic>MW: Exalogic Core
 

In this Document

Purpose

Scope

Details

Exalogic Control vServer Distribution

Exalogic Control vServer Identification

Stopping and Starting the Entire Exalogic Control stack

Stopping the whole Exalogic Control stack

Starting the whole Exalogic Control stack

References

Applies to:

Purpose

This document describes how to stop and start the entire Exalogic Control stack in EECS v2.0.6.0.0.

Scope

Details

Exalogic Control vServer Distribution

The number of VMs that make up the Exalogic Control stack has been reduced from five in EECS 2.0.4.0.x to three in 2.0.6.0.0

The Exalogic Control stack in EECS 2.0.6.0.0 consists of the following VMs:

• Two Proxy Controller VMs : PC2 & PC1
• Exalogic Control VM: This VM hosts the Enterprise Controller (EC1), Oracle VM Manager (OVMM), and Oracle Virtual Assembly Builder (OVAB) components. The same VM also hosts the Oracle database (DB) used by EM Ops Center and Oracle VM Manager. In earlier releases, each of these components was hosted on a dedicated VM.

Exalogic Control vServer Identification

Exalogic Control vServers can be identified as follows. On the master compute node (first node of the rack), execute the following command to identify the path to the vm.cfg file for each of the Exalogic Control Stack core vServers:

1. Fresh EECS 2.0.6.0.0 install:
   
   root@elorl03cn01 ~]# find /OVS -wholename '*VirtualMachines/*/vm.cfg' -exec grep -H 'simple_' {} \;|grep Exalogic
   /OVS/Repositories/0004fb0000030000209678530ae89eb7/VirtualMachines/0004fb000006000079ec6d30f251addb/vm.cfg:OVM_simple_name = 'ExalogicControl'
   /OVS/Repositories/0004fb0000030000209678530ae89eb7/VirtualMachines/0004fb0000060000f18dd947707040f0/vm.cfg:OVM_simple_name = 'ExalogicControlOpsCenterPC1'
   /OVS/Repositories/0004fb0000030000209678530ae89eb7/VirtualMachines/0004fb0000060000ca43c340c7c165d7/vm.cfg:OVM_simple_name = 'ExalogicControlOpsCenterPC2'
   root@elorl03cn01 ~]#
     
   
   
2. Upgraded (from EECS 2.0.4.x ) EECS 2.0.6.0.0 install:
   
   [root@tlv04cn01 ~]# find /OVS -wholename '*VirtualMachines/*/vm.cfg' -exec grep -H 'simple_' {} \;|grep Exalogic
   /OVS/Repositories/0004fb000003000046192057794d6d5f/VirtualMachines/0004fb00000600002089fc94fb7e61f0/vm.cfg:OVM_simple_name = 'ExalogicControlOpsCenterPC1'
   /OVS/Repositories/0004fb000003000046192057794d6d5f/VirtualMachines/0004fb00000600002bc872940f690a75/vm.cfg:OVM_simple_name = 'ExalogicControlOpsCenterPC2'
   /OVS/Repositories/0004fb000003000046192057794d6d5f/VirtualMachines/0004fb0000060000403b9aed6824cbb8/vm.cfg:OVM_simple_name = 'ExalogicControlOVMM'
   [root@tlv04cn01 ~]#
     
   Note both the full path and UUID of each vServer entry and for later use (the UUID will be the name of the parent directory containing the vm.cfg file).

Stopping and Starting the Entire Exalogic Control stack

Stopping the whole Exalogic Control stack

To stop the Exalogic Control stack, complete the following steps:

1. SSH to the PC1 vServer and run the following commands: 
   

   # /opt/sun/xvmoc/bin/proxyadm stop -w -v
   # shutdown -h now

   
     

   Note:
   
   Note the compute node (e.g. compute node 03) on which the PC1 vServer was running.   
   

2. SSH to the PC2 vServer and run the following command: 
   

   # /opt/sun/xvmoc/bin/proxyadm stop -w -v
   # shutdown -h now

   
     

   Note:
   
   Note the compute node (e.g. compute node 02) on which the PC2 vServer was running.   
   

3. SSH to the Exalogic Control vServer and run the following commands: 
   

   # /opt/sun/xvmoc/bin/ecadm stop -w -v
   # service ovmm stop
   # service oracle-db stop
   # shutdown -h now

    

   Note: Note the compute node (e.g. compute node 01) on which the Exalogic Control vServer was running with services: DB, EC1, OVAB, and OVMM

     
   

Starting the whole Exalogic Control stack

In order to start any of the Exalogic Control stack vServers, you need to connect the respective nodes hosting them.

Connect to the compute node that run each Exalogic Control vServer, then start each core vServer with the command below.

1. Start the Exalogic Control vServer. Ensure that it is started on the same node on which it was running earlier.

2. SSH to the Exalogic Control vServer and ensure that the required services are running:

   1. Ensure that the Oracle Database service is running:

      Note to Support: The command "service oracle-db status" do not work in 2.0.6.0.0. There is an Enhancement Request filed ( Bug 17827606/ Bug 17827529 ) to get this backported to 2.0.6.0.x. Once this is backported, the command will be available.
      
      Find out the status of the service: 
      

      # service oracle-db status
      
      Till the Enhancement is fixed, you can verify if the 'oracle-db' service configured to start runlevel or not and validate its process is running or not, as follows:
      
      
      [root@elorl03ec01-adm ~]# /sbin/chkconfig --list | grep oracle-db
      oracle-db          0:off    1:off    2:off    3:on    4:on    5:on    6:off
      [root@elorl03ec01-adm ~]# ps -ef | grep pmon
      oracle    8509     1  0 07:54 ?        00:00:00 ora_pmon_elctrldb
      root     10948 14119  0 10:09 pts/2    00:00:00 grep pmon
      [root@elorl03ec01-adm ~]#
        
      If the 'ora_pmon_elctrldb' process is not running, then start it with the next step.
      
      

      If the service is not running, run the following command: 
      

      # service oracle-db start

   2. Ensure that the OVMM (Oracle VM Manager) service is running:

      Find out the status of the service: 
      

      # service ovmm status

      If the service is not running, run the following command: 
      

      # service ovmm start

   3. Ensure that the OVAB (Oracle Virtual Assembly Builder) service is running. This step is applicable only to Exalogic version until Jan 2017 PSU (i.e. 2.0.6.2.170117) and lower, for latest Exalogic releases 2.0.6.2.170418 and 2.0.6.3.x, this step is not applicable.  

      Note:

      We will not be supporting Oracle Virtual Assembly Builder (OVAB) on the 2.0.6.3.0 Exalogic release or future releases of Exalogic Elastic Cloud Software. Even though we will continue to support customers who are running OVAB on Exalogic versions up until 2.0.6.2.170418, our recommendation to customers is not to adopt OVAB on Exalogic henceforth, if they desire to be aligned with the Exalogic Elastic Cloud Software Roadmap. Please reach out to Support if you are an existing Exalogic customer using OVAB and have questions on this roadmap.

      Please review Document ID : 1965704.1 - Direction For Oracle Virtual Assembly Builder (OVAB) On Exalogic  

      For Exalogic running on Jan 2017 PSU or lower , find out the status of the service on Exalogic version : 

      # service ovab status

      If the service is not running, run the following command: 
      

      # service ovab start

   4. Ensure that the EC1 (Enterprise Controller) service is running:

      Find out the status of the service: 
      

      # /opt/sun/xvmoc/bin/ecadm status

      If the service is offline, run the following command: 
      

      # /opt/sun/xvmoc/bin/ecadm start -w -v

3. Start the PC2 vServer. Ensure that it is started on the same node on which it was running earlier. PC2 can be started using xm create command as follows:

   xm create <path to vm.cfg of PC2 vserver>
4. SSH to the PC2 vServer and ensure that the PC2  service is running:

   Find out the status of the service: 
   

   # /opt/sun/xvmoc/bin/proxyadm status

   If the service is offline, run the following command: 
   

   # /opt/sun/xvmoc/bin/proxyadm start -w -v

5. Repeat the above steps 3 and 4 for PC1.

References

This solution has no attachment